A method and system for controlling vibration in a turning-milling process

Abstract

This invention discloses a vibration control method and system for milling and turning machining, relating to the field of milling and turning machining technology. The method includes: acquiring process signals associated with the position of the turning tool during historical turning operations of a previously machined workpiece; matching a corresponding turning tool working path for a newly machined workpiece based on the physical feature information, and associating the corresponding process signals according to the turning tool working path; matching a target spindle speed at each position on the turning tool working path based on the process signals; and controlling the newly machined workpiece to perform turning operations along the turning tool working path in the machine tool based on the target spindle speed. This invention can predict the actual cutting response characteristics at each point on the milling path in advance and adjust the milling spindle speed based on the prediction results, effectively improving the stability and surface quality of milling and turning machining.

Description

Technical Field

[0001] This invention relates to the field of milling and turning technology, and specifically to a vibration control method and system for milling and turning. Background Technology

[0002] In the field of milling and turning, especially for workpieces with low rigidity such as slender shafts or thin-walled parts, vibration generated during machining is a long-standing and difficult-to-eliminate technical problem. This vibration not only severely degrades the quality of the machined surface, leaving noticeable chatter marks, but also accelerates tool wear and, in extreme cases, damages the tool or workpiece. In multi-stage continuous machining scenarios, the interaction between processes further complicates the vibration problem. Preceding machining operations, such as turning, not only alter the workpiece geometry but also introduce residual stress in the surface and subsurface regions, potentially leading to work hardening of the localized material. This physical "imprint" left by the preceding operation changes the localized material mechanical properties of the workpiece surface, thus affecting the machining stability of subsequent operations (such as milling).

[0003] Existing vibration control methods typically assume that the workpiece material is homogeneous and isotropic, failing to adequately consider the localized and non-uniform changes in workpiece material properties caused by preceding machining processes. For example, in milling and turning of a rotating part with a large aspect ratio, the preceding turning operation leaves a residual stress field on the part surface that is non-uniformly distributed axially and is related to the tool state. This leads to unknown changes in the local material rigidity at subsequent milling points. This non-uniform local rigidity introduced by the preceding process alters the vibration characteristics of the milling system. The stable speed range calculated based on the assumption of homogeneous material has shifted its boundaries when facing a real surface that has undergone "work hardening" and "stress modification." A speed that was previously considered stable may fall into a newly generated unstable region due to residual stress, thus inducing chatter. Therefore, how to predict the actual cutting response characteristics of each point on the upcoming milling path without adding extra measurement steps or significantly affecting the machining cycle time, and adjust the milling spindle speed based on the prediction results, so as to actively avoid cutting chatter induced by local deviation of the system stability boundary, is a technical problem that urgently needs to be solved in the current turning and milling composite machining field. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this application provides a vibration control method and system for milling and turning, which has the advantages of being able to predict the actual cutting response characteristics of each point on the milling path in advance and adjust the milling spindle speed based on the prediction results, thereby actively avoiding cutting chatter induced by local shifts in the system stability boundary, and effectively improving the stability and surface quality of milling and turning.

[0005] This application provides a vibration control method for milling and turning processes, including:

[0006] Collect process signals associated with the position of the turning tool in the historical turning process of the machined workpiece, and establish the physical characteristic information of the machined workpiece under the turning process based on the process signals;

[0007] Based on physical feature information, the corresponding turning tool working path is matched for the new workpiece, and the corresponding process signal is associated with the turning tool working path.

[0008] The target spindle speed corresponding to each position on the working path of the turning tool is determined based on the process signal matching.

[0009] Based on the target spindle speed control, the newly machined workpiece is turned along the working path of the turning tool in the machine tool.

[0010] The above method can detect the actual cutting response characteristics of each point on the milling path in advance, and adjust the spindle speed accordingly, effectively avoiding cutting chatter caused by local rigidity changes due to previous machining, and improving machining stability.

[0011] Based on process signal matching, the target spindle speed corresponding to each position on the working path of the turning tool includes:

[0012] Preset adjustment rules are set based on the mapping relationship between process signals in physical feature information and the stable spindle speed of milling.

[0013] The target spindle speed at each position on the working path of the turning tool is determined based on preset adjustment rules.

[0014] By establishing a mapping relationship between process signals and stable spindle speed and setting adjustment rules, the determination of the target spindle speed becomes more accurate and intelligent.

[0015] The target spindle speed at each position on the working path of the turning tool is determined based on preset adjustment rules, including:

[0016] The process signals on the working path of the turning tool are analyzed to obtain a first characteristic index and a second characteristic index. The first characteristic index represents the impact characteristics of the process signals on the working path of the turning tool, and the second characteristic index represents the energy characteristics of the process signals on the working path of the turning tool.

[0017] The turning physical state of the newly machined workpiece is determined based on the first and second characteristic indicators.

[0018] Select the corresponding adjustment rule from the preset adjustment rules based on the physical state of the turning process;

[0019] The target spindle speed at different positions on the toolpath is determined based on the corresponding adjustment rules and the process signals on the working path of the turning tool.

[0020] The above scheme introduces first and second characteristic indicators to analyze process signals and judges the physical state of workpiece turning based on these indicators, thereby selecting adjustment rules more precisely and improving the adaptability of vibration control.

[0021] Determining the turning physical state of a newly machined workpiece based on the first and second characteristic indicators includes:

[0022] When the combination relationship between the first feature index and the second feature index satisfies the preset fuzzy physical state, the position on the working path of the turning tool is determined as the fuzzy state position.

[0023] Extract a path segment containing the ambiguous state position on the working path of the turning tool;

[0024] The first and second feature indices distributed on the path segment are analyzed, and the third feature indices representing the fuzzy physical state within the path segment are obtained.

[0025] The turning physical state of the fuzzy state is determined based on the characteristics of the third indicator. The turning physical state includes: composite physical layered structure and microscopic defects inside the workpiece material.

[0026] The above scheme addresses fuzzy physical states by extracting path segments and analyzing feature indicators, introducing a third indicator feature to more accurately determine the microscopic defects or complex physical layered structure of the workpiece, thereby improving the ability to identify complex workpiece states.

[0027] The first and second feature indices distributed along the path segment are analyzed, and the third feature indices representing the fuzzy physical state within the path segment are obtained, including:

[0028] For the sequence of the first or second characteristic index distributed on the path segment, a reference sequence is generated to represent the continuously changing part of the sequence. The reference sequence has been filtered out to remove local mutation information caused by micro defects.

[0029] Based on the benchmark sequence, a statistical value representing the spatial continuity of the benchmark sequence is determined as the third characteristic index.

[0030] By generating a baseline sequence and filtering out local abrupt changes caused by microscopic defects, the third characteristic index can more accurately reflect the spatial continuity of the sequence, thereby improving the characterization accuracy of the material's physical characteristics.

[0031] For a sequence of first or second characteristic indicators distributed along a path segment, generating a baseline sequence representing the continuously changing portion of the sequence includes:

[0032] Based on the data points in the sequence and the neighboring data points of the data points in the sequence, a local variation index is determined to represent the degree of local variation of the data points.

[0033] Based on the local variation index, data points in the sequence are distinguished into trend points representing continuous changes or abrupt changes caused by micro-defects;

[0034] A baseline sequence is generated by differentiating results based on trend points and abrupt change points. The value at the position corresponding to the trend point in the baseline sequence is determined by the value of the trend point in the sequence, and the value at the position corresponding to the abrupt change point in the baseline sequence is determined by the value of the trend point adjacent to the abrupt change point in the sequence.

[0035] The above scheme refines the generation process of the baseline sequence. By distinguishing between trend points and abrupt change points, it ensures that the baseline sequence can accurately represent the continuously changing parts and effectively remove noise interference.

[0036] The value at the position corresponding to the mutation point in the baseline sequence is determined by the values ​​of the trend points adjacent to the mutation point in the sequence, including:

[0037] In the sequence, identify a first trend point located before the mutation point and a second trend point located after the mutation point;

[0038] Based on the values ​​and positions of the first and second trend points in the sequence, the value of the mutation point at the corresponding position in the baseline sequence is determined.

[0039] The above scheme provides a specific method for determining the corresponding value of mutation points in the baseline sequence, and further optimizes the construction accuracy of the baseline sequence by interpolating with nearby trend points.

[0040] Determining the turning physical state of a newly machined workpiece based on the first and second characteristic indicators includes:

[0041] A two-dimensional feature space is constructed using the first and second feature indices as coordinate axes. The two-dimensional feature space is used to delineate the corresponding discrimination regions for different turning physical states.

[0042] The first feature index and the second feature index at a position along the working path of the turning tool constitute a feature point located in a two-dimensional feature space.

[0043] The turning physical state of the newly machined workpiece is determined based on the discrimination region in which the feature points fall in the two-dimensional feature space.

[0044] The above scheme, by constructing a two-dimensional feature space and delineating the discrimination region, enables intuitive and quantitative judgment of the physical state of turning, simplifying the identification process of complex states.

[0045] The target spindle speed at each position on the working path of the turning tool is determined based on preset adjustment rules, including:

[0046] Determine the initial target spindle speed based on the process signal characteristics of the working path distribution of the turning tool;

[0047] During the subsequent turning operations of the newly machined workpiece, real-time process signals reflecting the current interaction state between the turning tool and the newly machined workpiece are collected.

[0048] The characteristics of the current wear state of the turning tool are determined from the collected real-time process signals;

[0049] The compensation logic is set according to the characteristics of the current wear state of the turning tool, and the spindle speed compensation amount is determined based on the compensation logic;

[0050] The target spindle speed is obtained by correcting the initial target spindle speed based on the spindle speed compensation amount.

[0051] The above scheme introduces real-time process signals and tool wear characteristics based on the initial target spindle speed, and dynamically corrects the spindle speed through compensation logic, thereby improving the real-time performance and adaptability of vibration control.

[0052] This application also proposes a milling and turning vibration control system for executing the above-described milling and turning vibration control method, the system comprising:

[0053] The feature construction module is used to collect process signals associated with the position of the turning tool in the historical turning process of the workpiece, and to establish the physical feature information of the workpiece under the turning process based on the process signals.

[0054] The path construction module is used to match the corresponding turning tool working path for the new workpiece based on physical feature information, and to associate the corresponding process signal according to the turning tool working path.

[0055] The spindle speed determination module is used to match the process signal to determine the target spindle speed at each position on the working path of the turning tool.

[0056] The machine tool control module is used to control the newly machined workpiece to perform turning operations along the working path of the turning tool in the machine tool based on the target spindle speed.

[0057] In summary, the turning and milling vibration control method and system disclosed in this application establishes workpiece physical characteristic information by collecting process signals, and matches the tool working path and process signals to the new workpiece based on this information. This allows for the matching and control of the target spindle speed at each position, thereby enabling the prediction of the actual cutting response characteristics at each point on the milling path in advance. Based on this prediction, the spindle speed is adjusted, thus actively avoiding cutting chatter induced by local shifts in the system stability boundary. This effectively improves the stability and surface quality of turning and milling. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is a flowchart of the vibration control method for milling and turning in an embodiment of the present invention;

[0060] Figure 2 This is a flowchart illustrating how the target spindle speed at each position on the working path of a turning tool is determined based on preset adjustment rules in an embodiment of the present invention.

[0061] Figure 3 This is a schematic diagram of the vibration control system for milling and turning in an embodiment of the present invention. Detailed Implementation

[0062] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0063] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0064] This invention indirectly obtains the physical characteristics of a workpiece after turning by utilizing existing historical machining data, and uses this information to predict the response of subsequent machining. By analyzing process signals related to the turning tool position in the historical turning operations of the workpiece, a physical characteristic information model of the workpiece during the turning process can be established. Based on this physical characteristic information, a corresponding turning tool working path can be matched for the new workpiece, and the corresponding process signals can be associated. Using these process signals, the target spindle speed corresponding to each position on the turning tool working path can be matched. Finally, based on these target spindle speeds, the new workpiece is controlled to perform turning operations along the turning tool working path in the machine tool, thereby achieving vibration control.

[0065] Specifically, Figure 1 A flowchart of a vibration control method for milling and turning machining according to an embodiment of the present invention is shown, which specifically includes:

[0066] S101. Collect process signals associated with the position of the turning tool in the historical turning process of the processed workpiece, and establish physical characteristic information of the processed workpiece under the turning process based on the process signals.

[0067] It should be noted that the process signals here refer to the physical quantities or state quantities acquired in real time during the turning process, corresponding to the specific position of the turning tool on the workpiece. These can be acquired using various sensor technologies, such as force sensors to acquire cutting force signals, accelerometers to acquire vibration signals, acoustic emission sensors to acquire acoustic emission signals, or current sensors to acquire the current signals of the spindle or feed axis. The purpose is to capture and quantify the dynamic response of the workpiece material to the tool action during turning, providing raw data for establishing subsequent physical characteristic information. Physical characteristic information refers to the data set or model constructed by analyzing, processing, and modeling the process signals, which characterizes the mechanical properties or states of the machined workpiece under turning action, such as its internal material structure, residual stress distribution, and local hardness changes. It can be represented in the form of data tables, feature vectors, machine learning models, or statistical models. Its purpose is to transform the complex machining history into a quantifiable description of the workpiece's intrinsic properties that can be used to predict subsequent machining behavior.

[0068] S102. Match the corresponding turning tool working path to the new workpiece based on the physical feature information, and associate the corresponding process signal according to the turning tool working path.

[0069] It should be noted that, based on the geometry, material type, and preset machining path of the new workpiece, this method searches for or derives data or models from the established physical feature information that are closest to the physical response or state that the new workpiece may produce on a specific turning tool working path. This can be achieved using similarity calculation, pattern recognition algorithms, or rule-based reasoning systems. The purpose is to effectively transfer historical experience knowledge to the machining prediction of the new workpiece, thereby predicting the machining characteristics of the new workpiece on a specific path.

[0070] S103. Based on the process signal, match the target spindle speed corresponding to each position on the working path of the turning tool;

[0071] The target spindle speed here refers to the spindle rotation speed determined by calculation or query for each specific position on the working path of the turning tool, based on the physical state of the workpiece reflected by the process signal associated with that position. This speed effectively suppresses vibration and optimizes machining performance. It can be generated using a preset lookup table, dynamic optimization algorithm, or adaptive control strategy based on real-time feedback. The purpose is to achieve dynamic adjustment of the spindle speed to adapt to the local machining characteristics of the workpiece changing along the path, thereby avoiding cutting chatter.

[0072] First, to obtain the inherent physical characteristics of the workpiece during turning, this method collects process signals closely related to the turning tool position in historical turning operations of the machined workpiece. These signals, such as cutting forces, vibrations, or acoustic emissions, objectively reflect the dynamic response when the tool interacts with the workpiece material, containing key information such as the local rigidity of the workpiece material, residual stress distribution, and work hardening. Based on these rich process signals, this method further establishes the physical characteristic information of the machined workpiece during the turning process, which is equivalent to constructing a knowledge base or model that can describe the "machining imprint" of the workpiece. On this basis, when a new workpiece needs to be machined, this method matches the corresponding turning tool working path for the new workpiece based on the established physical characteristic information. This means that the system can predict the behavior of the new workpiece on a specific tool path based on its machining requirements and its potential "machining imprint." Once the tool working path is matched, the system can associate the corresponding process signals on that path, which are predictions of the possible physical responses of the new workpiece at the corresponding positions based on historical data. Subsequently, this method utilizes these predicted process signals to precisely match the target spindle speed corresponding to each position on the turning tool's working path. This step is crucial for achieving dynamic vibration control, as it allows the system to intelligently adjust the spindle speed based on the changing local physical state of the workpiece along the path, rather than using fixed parameters that may cause vibration. This dynamic adjustment capability based on predicted signals enables the system to proactively avoid unstable cutting zones. Finally, based on these optimized target spindle speeds, this method precisely controls the newly machined workpiece to perform turning operations along the preset turning tool's working path on the machine tool. In this way, the machine tool can adjust the spindle speed in real time according to the actual local characteristics of the workpiece, thereby effectively suppressing vibrations that may occur during machining and ensuring machining quality and efficiency. The entire process forms a closed-loop prediction-adjustment-control mechanism, enabling the machining process to adapt to non-uniform variations in the workpiece material.

[0073] S104. Based on the target spindle speed, control the newly machined workpiece to perform turning operations along the working path of the turning tool in the machine tool.

[0074] In practice, during historical turning operations, three-axis accelerometers and force sensors mounted on the machine tool spindle box or tool holder can be used to collect vibration and cutting force signals at different positions as the turning tool moves along the workpiece axis in real time. These signals are recorded synchronously with the tool's instantaneous position data, forming time-series data. Subsequently, these process signals can be input into a data processing unit, which can be a high-performance industrial computer running data analysis algorithms. This algorithm extracts features from the signals, such as calculating the root mean square value, peak factor, and spectral characteristics, and associates these features with the tool position to construct a multi-dimensional physical feature information database. This database can be stored as a mapping table containing the correlation between different workpiece types, machining parameters, and corresponding physical features. When a new workpiece needs to be machined, its preset turning tool working path information is input into the path construction module. This module queries the aforementioned physical feature information database. Based on the new workpiece's material, geometry, and expected machining path, it uses pattern recognition or similarity matching algorithms to match the most similar historical data segments from the database to the new workpiece's path and associates them with the predicted process signals corresponding to each position on that path. For example, if a path segment of the new workpiece is highly similar in material and geometry to a specific path segment of a machined workpiece in the historical database, the process signal features of that historical path segment will be associated with the corresponding position on the new workpiece. Next, the spindle speed determination module receives these predicted process signals. This module can have a built-in preset adjustment rule model, such as a mapping model trained based on fuzzy logic or a neural network, which maps the features of the process signals to the optimal spindle speed. For instance, if a high vibration risk is predicted for a certain position, the model will output a lower spindle speed or a spindle speed that avoids the resonance zone. This model is trained offline based on a large amount of historical machining data and stability maps. Finally, the machine tool control module receives these target spindle speed sequences determined for each path position. During actual turning operations, the machine tool's CNC system dynamically reads from the sequence based on the real-time position of the tool and instructs the spindle to rotate at the corresponding target speed. For example, when the tool moves to a specific point on the path, the CNC system immediately adjusts the spindle speed to the preset target spindle speed for that point, thereby ensuring that the spindle speed can adaptively adjust according to changes in the local physical properties of the workpiece throughout the turning process, effectively suppressing vibration.

[0075] The method in this embodiment of the invention combines process signals associated with the turning tool position collected in historical turning operations with workpiece physical feature information established based on these signals in a data-driven and predictive manner. This solves the problem of non-uniform residual stress fields left on the workpiece surface by previous turning operations, which leads to unknown changes in local material rigidity at subsequent machining points and easily induces cutting chatter. This achieves the effects of improving machining stability, suppressing vibration, and optimizing machining quality. The method in this embodiment of the invention effectively solves the problem of cutting chatter easily induced in mill-turn machining when the residual stress field left on the workpiece surface by previous turning operations is non-uniformly distributed along the axial direction and is related to the turning tool state, leading to unknown changes in local material rigidity at subsequent machining points. This method, through in-depth mining of historical turning process signals and the establishment of physical feature information, enables the advance prediction of the actual cutting response characteristics of the newly machined workpiece at each point on the turning tool's working path. This data-driven predictive capability allows the system to match the optimal target spindle speed for each position on the tool path without adding extra measurement steps or significantly affecting the machining cycle time. Therefore, this method can actively avoid cutting chatter induced by local shifts in the system stability boundary, significantly improve the stability of the machining process, improve the quality of the machined surface, and extend the tool life.

[0076] It should be noted that the steps for matching the target spindle speed at each position on the working path of the turning tool based on the process signal include: setting a preset adjustment rule based on the mapping relationship between the process signal in the physical feature information and the stable spindle speed of milling; and determining the target spindle speed at each position on the working path of the turning tool based on the preset adjustment rule.

[0077] It should be noted that the mapping relationship between the process signal in the physical feature information and the stable spindle speed in milling refers to establishing the correspondence between the process signal and the spindle speed range that can maintain system stability and avoid chatter during milling. This can be achieved by means of data fitting, machine learning model training or expert system rule base, etc. The purpose is to transform complex machining state information into an operable basis for speed adjustment.

[0078] It should be noted that the preset adjustment rules refer to a series of strategies or algorithms pre-set according to the above mapping relationship to guide the adjustment of the spindle speed. They can be implemented by means of conditional judgment statements, decision trees, fuzzy logic rule sets or prediction models based on neural networks, etc. The purpose is to ensure that the adjustment of the spindle speed is predictable and stable.

[0079] The method in this embodiment of the invention establishes a mapping relationship between process signals and stable milling spindle speeds, and sets preset adjustment rules based on this, thereby transforming abstract process signals into specific speed adjustment strategies. During actual machining, for each point on the path, the system can dynamically determine a target spindle speed that matches the actual machining state of that point (reflected by the process signal). This dynamic adjustment capability means that the spindle speed is no longer fixed, but can respond in real time to changes in workpiece material properties, tool wear, or other machining conditions, thus avoiding vibrations caused by parameter mismatches. This method can effectively address the non-uniform residual stress field and local material rigidity changes introduced on the workpiece surface by preceding turning operations. Therefore, without adding extra measurement steps or significantly affecting the machining cycle time, it can pre-detect the actual cutting response characteristics of each point on the upcoming milling path, and adjust the milling spindle speed based on this detection result, actively avoiding cutting chatter induced by local shifts in the system stability boundary.

[0080] When determining the target spindle speed at each position along the turning tool's working path, the system can acquire process signals at the current position in real time. For example, vibration signals can be obtained through an accelerometer mounted on the tool holder, or cutting power signals can be obtained through a spindle power sensor. These real-time signals, after preprocessing and feature extraction, are fed as input to a pre-trained mapping model. The model outputs a suggested stable spindle speed range. Then, according to preset adjustment rules, a specific target spindle speed is selected from this suggested range. For example, if the rules stipulate that a lower stable speed is preferred in the roughing stage to ensure cutting stability, while a higher stable speed is preferred in the finishing stage to improve surface quality, the system will determine the final target spindle speed based on the current process type and the range output by the model. In this way, the spindle speed adjustment can dynamically adapt to the actual machining state of the workpiece and the preset optimization goals.

[0081] Specifically, Figure 2 This invention illustrates a flowchart illustrating how, in an embodiment of the present invention, the target spindle speed at each position on the working path of a turning tool is determined based on preset adjustment rules. The flowchart specifically includes:

[0082] S201. Analyze the process signals on the working path of the turning tool to obtain the first characteristic index and the second characteristic index;

[0083] It should be noted that the first characteristic index here represents the impact characteristics of the process signal along the working path of the turning tool, while the second characteristic index represents the energy characteristics of the process signal along the working path of the turning tool. The first characteristic index is a quantitative value representing the impact characteristics of the process signal obtained after analyzing the process signal along the working path of the turning tool. Specifically, it can be obtained by extracting and calculating the instantaneous peak value, peak factor, or high-frequency components of the process signal. Its purpose is to reflect the magnitude and frequency of the transient impact load experienced by the tool when it contacts the workpiece during the cutting process, such as the impact response caused by hard spots, inclusions, or microcracks on the workpiece surface. The second characteristic index is a quantitative value representing the energy characteristics of the process signal obtained after analyzing the process signal along the working path of the turning tool. Specifically, it can be obtained by calculating the root mean square value, total energy, or energy in a specific frequency band of the process signal. Its purpose is to reflect the overall stability of the cutting process, the magnitude of the cutting force, and energy consumption, such as energy dissipation from friction and plastic deformation during the cutting process.

[0084] S202. Determine the turning physical state of the newly machined workpiece based on the first and second characteristic indicators;

[0085] It should be noted that the turning physical state refers to the classification of the material mechanical properties or surface integrity state of a newly machined workpiece after turning, based on the combination of the first and second characteristic indices. Specifically, it can be a comprehensive characterization of the residual stress distribution, local hardness changes, micro-defects, or work hardening degree on the workpiece surface. Its purpose is to identify non-uniform or abnormal areas that may exist in the workpiece after the turning process, which may affect the stability of subsequent milling processes.

[0086] When determining the turning physical state of a newly machined workpiece based on the first and second characteristic indices, a multi-dimensional feature space can be constructed, with the first and second characteristic indices serving as coordinate axes. Beforehand, using extensive experimental or simulation data, different turning physical states, such as residual compressive stress zones, residual tensile stress zones, micro-defect zones, and homogeneous material zones, are delineated in the feature space as corresponding discrimination regions. Once the new first and second characteristic indices are calculated, they constitute a point in the feature space. By determining which discrimination region this point falls into, the turning physical state of the current workpiece position can be determined. For example, high impact characteristics and medium energy characteristics may indicate the presence of micro-defects in the workpiece, while low impact characteristics and high energy characteristics may indicate homogeneous workpiece material and smooth cutting.

[0087] S203. Select the corresponding adjustment rule from the preset adjustment rules according to the physical state of turning;

[0088] Here, selecting the corresponding adjustment rule from the preset adjustment rules means selecting the most suitable specific adjustment strategy for the current physical state of the workpiece from a set of pre-established spindle speed adjustment strategies set for different physical states, based on the determined turning physical state. Specifically, this can be done by consulting a preset rule base, decision tree, or matching based on a machine learning model. The purpose is to ensure that the subsequent target spindle speed setting can accurately adapt to the actual local characteristics of the workpiece, thereby optimizing the stability of milling.

[0089] The adjustment rules selected here are combined with process signals on the turning tool path to determine the target spindle speed at different positions along the tool path. This combination makes the spindle speed adjustment no longer a simple, static process, but rather a dynamic and adaptive correction based on differences in the local physical state of the workpiece. For example, when a hard spot or high residual stress is detected in a certain area of ​​the workpiece, the system can select an adjustment rule aimed at reducing impact or improving stability, thereby appropriately reducing the spindle speed in that area; conversely, if the workpiece is identified as being in good condition, a rule that allows for higher machining efficiency may be selected. It is precisely because of this rule selection mechanism based on actual physical state that the setting of the target spindle speed can more accurately adapt to the local characteristics of the workpiece, effectively avoiding machining instability caused by workpiece non-uniformity, thus optimizing machining efficiency while ensuring machining quality.

[0090] Specifically, when selecting the corresponding adjustment rule from preset adjustment rules based on the turning physical state, a rule base can be established. This rule base contains a series of predefined adjustment rules, each associated with one or more specific turning physical states. For example, for a physical state identified as a "microscopic defect area," an adjustment rule of "reducing the spindle speed and decreasing the feed rate" can be preset; for a physical state identified as a "high residual compressive stress area," an adjustment rule of "maintaining the current spindle speed but slightly increasing the feed rate to promote stress release" can be preset. Once the system determines the physical state of the workpiece, it directly retrieves and selects the adjustment rule corresponding to that state from the rule base.

[0091] S204. Determine the target spindle speed at different positions on the toolpath based on the corresponding adjustment rules and process signals on the working path of the turning tool.

[0092] Here, when determining the target spindle speed at different positions on the toolpath based on the selected adjustment rule and the process signals on the turning tool's working path, the selected adjustment rule can be a function or algorithm. It takes the real-time characteristics of the process signals as input and outputs a correction amount or target value for the spindle speed according to its internal logic. For example, if the selected adjustment rule is "reduce spindle speed," a negative speed compensation amount can be dynamically calculated based on the impact intensity of the process signals and superimposed on the initially set spindle speed to obtain the final target spindle speed at that position. This dynamic adjustment mechanism ensures that the spindle speed can accurately respond to local physical changes in the workpiece, thereby optimizing the stability of the machining process.

[0093] The method in this embodiment of the invention analyzes the process signals along the working path of the turning tool to obtain a first characteristic index and a second characteristic index representing impact and energy characteristics, which can accurately determine the turning physical state of the newly machined workpiece. Based on the identification of the actual physical state of the workpiece, the most suitable adjustment rule for the current workpiece characteristics can be selected from preset adjustment rules, thereby achieving targeted and adaptive adjustment of the target spindle speed. Therefore, the setting of the spindle speed is no longer static, but can dynamically adapt to changes in the local material properties of the workpiece, effectively avoiding cutting chatter caused by uneven residual stress distribution or microscopic defects, significantly improving the stability and quality of the machining process, and optimizing machining efficiency.

[0094] It should be noted that the steps for determining the turning physical state of the newly machined workpiece based on the first and second characteristic indicators include: when the combination relationship between the first and second characteristic indicators satisfies the preset fuzzy physical state, the position on the working path of the turning tool is determined as the fuzzy state position; a path segment containing the fuzzy state position is extracted on the working path of the turning tool; the first and second characteristic indicators distributed on the path segment are analyzed, and a third indicator feature representing the fuzzy physical state within the path segment is obtained; the turning physical state of the fuzzy state position is determined based on the third indicator feature, and the turning physical state includes: composite physical layered structure and microscopic defects inside the workpiece material.

[0095] The fuzzy physical state, as defined here, refers to a combination of the first and second feature indices that cannot be clearly categorized into a known, clear turning physical state. For example, the combined value may fall at the boundary of multiple discrimination regions or in an undefined area. This can be achieved using fuzzy logic rules, probabilistic models, or low-confidence outputs of machine learning classifiers. Its purpose is to identify the uncertainty in the current state judgment and avoid direct misclassification. The fuzzy state position refers to a point on the turning tool's working path where the combination of the first and second feature indices is judged to satisfy the preset fuzzy physical state. This can be a discrete point or a very small region, its purpose being to accurately mark specific machining areas requiring further analysis. The path segment refers to a continuous path segment on the turning tool's working path, intercepted around one or more fuzzy state positions. This can be a fixed-length interval or a length adaptively determined based on signal characteristics, its purpose being to provide contextual information around the fuzzy state position for comprehensive analysis. The third indicator feature refers to one or a set of quantitative indicators extracted from the first and second characteristic indicators distributed along the path segment. These indicators characterize the fuzzy physical state properties within that path segment and can reflect the trend, volatility, frequency components, or specific patterns of the indicator sequence. Their purpose is to extract key information from regional data to distinguish different fuzzy physical states. The composite physical layered structure refers to the multi-layered structure of physical properties formed on the surface or subsurface of the workpiece material during turning due to the combined effects of cutting force, thermal effects, and material deformation. This can manifest as hardness gradients, residual stress distribution, or grain structure changes. Its purpose is to identify changes in material properties caused by processing history. Microscopic defects within the workpiece material refer to small, discontinuous internal imperfections that exist before manufacturing or processing, or are newly generated during processing. These can include inclusions, pores, microcracks, or grain boundary defects. Their purpose is to identify inherent material abnormalities or incidental local anomalies during processing.

[0096] The method in this embodiment of the invention solves the problem of misjudgment that may occur when judging the physical state of turning by traditional methods by introducing the identification and in-depth analysis of fuzzy physical states. When the system initially judges that the combination of the first and second feature indices is in a preset fuzzy physical state, it no longer directly performs uncertain state classification, but first determines the current position on the working path of the turning tool as the fuzzy state position. This step avoids making wrong decisions based on ambiguous information and lays the foundation for subsequent accurate judgment. Subsequently, the system will extract a path segment on the working path of the turning tool that contains the fuzzy state position. The purpose of this operation is to expand a single fuzzy point into a region with contextual information, because the index at a single position may be affected by instantaneous noise or local anomalies, while the index distribution within a path segment can stably and comprehensively reflect the true characteristics of the material. Next, the system will analyze the first and second feature indices distributed on the path segment and obtain the third index feature representing the fuzzy physical state within the path segment. This process is an in-depth mining of signal features in a local area. For example, by analyzing the continuous change trend, fluctuation frequency, or specific pattern of the index, key information that distinguishes different fuzzy states can be extracted. This information cannot be provided by the index at a single position. Ultimately, based on the acquired third indicator features, the system can accurately determine the turning physical state of the fuzzy state location and clearly distinguish whether it is a composite physical layered structure or a microscopic defect inside the workpiece material.

[0097] This series of steps is closely integrated with the preceding steps, which determine the turning physical state using first and second characteristic indicators and select adjustment rules accordingly to determine the target spindle speed. The introduction of this solution enables precise and accurate physical state judgments when facing complex or uncertain material properties. By accurately identifying and classifying fuzzy states, it ensures that the subsequent selection of adjustment rules better matches the actual workpiece material properties, thereby ensuring accurate determination of the target spindle speed, avoiding machining vibrations caused by misjudgments, and ultimately improving machining quality and efficiency. Specifically, when the combination of the first and second characteristic indicators is fuzzy and it is difficult to directly determine the turning physical state, the method in this invention identifies the location of the fuzzy state and extracts a discriminative third indicator feature through in-depth analysis of the path segment where that location is located. Based on this third indicator feature, it is possible to accurately determine whether the fuzzy state location is caused by a complex physical layered structure or microscopic defects within the workpiece material. This avoids potential misjudgments that may occur with traditional methods in fuzzy states, thus providing a reliable basis for precise adjustment of the subsequent spindle speed, helping to improve machining quality, reduce vibration, and increase machining efficiency.

[0098] The first or second feature index distributed on the path segment in the embodiments of the present invention, and the acquisition of the third feature index representing the fuzzy physical state within the path segment, includes: generating a reference sequence representing the continuously changing part of the sequence for the sequence of the first or second feature index distributed on the path segment, wherein the reference sequence has filtered out local mutation information caused by microscopic defects; and determining a statistical value representing the spatial continuity of the reference sequence as the third feature index based on the reference sequence.

[0099] The reference sequence refers to the data sequence extracted from the original feature index sequence, representing its main continuous trend. It has been processed to remove local spikes or abrupt changes caused by discontinuous factors such as microscopic defects. The reference sequence can be generated using various signal processing techniques, such as applying low-pass filters, median filters, moving average filters, or wavelet transform-based denoising methods to smooth the original sequence and highlight its inherent continuous changes. Its purpose is to provide a processed data foundation that better reflects the true trend, facilitating subsequent quantification and judgment of fuzzy physical states. Spatial continuity refers to the smoothness or consistency of the reference sequence's changes along the path segment. It reflects the distribution characteristics of fuzzy physical states within the workpiece material. For example, if the fuzzy physical state is a composite physical layered structure, its corresponding reference sequence may change relatively smoothly, exhibiting high spatial continuity; if the fuzzy physical state is a microscopic defect within the workpiece material, its corresponding reference sequence may change relatively drastically, exhibiting low spatial continuity. Its purpose is to provide a criterion for distinguishing different types of fuzzy physical states by quantifying this continuity. In this context, statistical values ​​refer to one or more numerical values ​​used to quantify the spatial continuity of a benchmark sequence. These statistical values ​​may include, but are not limited to, the variance, standard deviation, mean absolute difference, and autocorrelation coefficient of the first or second derivative of the benchmark sequence, or a measure based on the fractal dimension of the sequence. The purpose is to transform abstract spatial continuity into calculable and comparable numerical values, thereby serving as a third characteristic indicator for subsequent judgment of fuzzy physical states.

[0100] The proposed solution processes the sequences of first or second characteristic indices distributed along a path segment to more accurately obtain a third indicator feature representing the fuzzy physical state within that path segment. Specifically, firstly, a baseline sequence is generated for the original first or second characteristic indicator sequences. This baseline sequence generation process is crucial; it is specifically designed to filter out local abrupt changes caused by microscopic defects within the workpiece material. These local abrupt changes manifest as spikes or outliers in the original sequence; they are not a true reflection of the overall physical state of the workpiece, but rather local, discrete interference. By generating the baseline sequence, this solution effectively "smooths out" these interferences, resulting in a processed data stream that better characterizes the continuously changing parts of the sequence. Based on this filtered-out baseline sequence, a statistical value representing its spatial continuity is determined as the third characteristic indicator. Spatial continuity is a property that measures the smoothness of sequence changes. When the fuzzy physical state is a complex physical hierarchical structure, its corresponding characteristic indicator sequence, after filtering out abrupt changes, usually exhibits a relatively smooth and continuous trend, thus its baseline sequence has high spatial continuity. Conversely, when the fuzzy physical state is a microscopic defect within the workpiece material, even after initial filtering, its baseline sequence may still exhibit some discontinuity or more severe fluctuations locally, resulting in low spatial continuity. By quantifying this spatial continuity, this scheme provides a quantified third feature index, making subsequent judgments based on the third index feature regarding the turning physical state of the fuzzy state (e.g., distinguishing between complex physical layered structures or microscopic defects within the workpiece material) more accurate. The introduction of this scheme overcomes the limitation of traditional methods that are easily affected by microscopic defects when directly analyzing the original feature index sequence when judging the turning physical state of a newly machined workpiece. By providing a more representative third index feature, this scheme improves the accuracy of classifying and judging fuzzy physical states. This improved accuracy further optimizes the subsequent selection of adjustment rules based on the turning physical state and the determination of the target spindle speed, thereby enhancing the robustness and effectiveness of the milling and turning vibration control method overall. This ensures stable machining even under complex workpiece material conditions and avoids machining vibration or quality problems caused by misjudgment.

[0101] When analyzing the first and second feature indices on a path segment containing fuzzy state locations, local abrupt changes caused by microscopic defects within the workpiece material can be effectively filtered out. This makes the acquired third feature indices representing the fuzzy physical state within the path segment more accurate, thereby improving the accuracy of fuzzy physical state judgment, avoiding misjudgments caused by interference from microscopic defects, and ensuring the accuracy of subsequent turning physical state judgment, ultimately improving the overall effect of vibration control in milling and turning processes.

[0102] The method in this embodiment of the invention generates a benchmark sequence representing the continuously changing portion of a sequence of first or second characteristic indicators distributed on a path segment. The steps include: determining a local variation index representing the degree of local variation of data points based on data points in the sequence and neighboring data points in the sequence; classifying data points in the sequence into trend points representing continuously changing portions or abrupt changes caused by microscopic defects based on the local variation index; and generating a benchmark sequence based on the distinction between trend points and abrupt changes, wherein the value corresponding to the trend point in the benchmark sequence is determined by the value of the trend point in the sequence, and the value corresponding to the abrupt change in the benchmark sequence is determined by the value of the trend point adjacent to the abrupt change in the sequence.

[0103] It should be noted that local variation indices are numerical values ​​used to quantify the degree of change of a data point in a sequence relative to its surrounding data points. They can be calculated by measuring the difference, slope, or statistical variance between the data point and its neighbors, aiming to identify potential abnormal fluctuations or stable regions within the sequence. Neighboring data points refer to the set of data points in the sequence that are geographically adjacent to the current data point or within a certain window. Specifically, this can include the preceding and following data points, or data points in a subsequence formed by extending several data points forward and backward from the current data point. Their purpose is to provide contextual information for assessing the local characteristics of the current data point. Trend points, on the other hand, are data points in the sequence with relatively small local variation indices, reflecting the overall or local continuous trend of change in the sequence. Specifically, a threshold can be set to identify data points with local variation indices below that threshold as trend points, aiming to characterize the stable or gradually changing parts of the sequence. A mutation point refers to a data point in the sequence with a large local variation index, caused by discontinuous factors such as microscopic defects. Specifically, it can be identified by setting a threshold, and data points with local variation indices exceeding that threshold are identified as mutation points. The purpose is to identify and distinguish instantaneous changes in the sequence caused by noise or abnormal events. A nearby trend point refers to a data point in the sequence that is close to the mutation point and has already been identified as a trend point. Specifically, it can be determined by searching for the nearest trend point in the directions before and after the mutation point. The purpose is to provide a reasonable alternative value based on the continuous trend of the sequence for the mutation point, thereby eliminating the influence of the mutation point.

[0104] The method in this embodiment of the invention refines the original sequence of the first or second feature indicators distributed along the path segment to generate a baseline sequence representing the continuously changing portion of the sequence. Specifically, firstly, for each data point in the sequence, a local variation index is calculated by combining information from its neighboring data points. This index quantifies the degree of change of the data point relative to its surrounding environment, thus providing a basis for subsequent classification. It is precisely because of the precise quantification of the degree of local variation that the system can effectively distinguish data points of different natures in the sequence. Secondly, based on the determined local variation index, the system intelligently divides the data points in the original sequence into two categories: one is trend points reflecting the stable or gradual trend of the sequence, and the other is abrupt change points caused by discontinuous factors such as microscopic defects. This distinction mechanism is the key to the accurate identification and isolation of abnormal information in this scheme. Finally, when generating the baseline sequence, for data identified as trend points, their values ​​in the baseline sequence directly adopt their original values, thus preserving the true continuous change information of the sequence; while for data identified as abrupt change points, their values ​​in the baseline sequence are determined by the values ​​of their neighboring trend points. This trend-point-based filling strategy effectively filters out local abrupt changes caused by microscopic defects, enabling the generated baseline sequence to more accurately represent the inherent continuous variation law of the sequence. Through the above processing, this scheme provides a purer and more representative baseline sequence. This baseline sequence is then used to determine the third index feature representing the fuzzy physical state within the path segment. As mentioned in the previous scheme, a statistical value representing the spatial continuity of the baseline sequence is determined as the third feature index based on the baseline sequence. Since the baseline sequence has effectively filtered out the interference caused by microscopic defects, the determined third index feature can more accurately reflect the true physical state of the workpiece material. This allows for higher accuracy and reliability in determining the turning physical state at the location of the fuzzy state, thus providing a more solid data foundation for subsequent adjustment of the spindle speed based on the physical state, ultimately improving the overall effect of vibration control in milling and turning.

[0105] In some preferred embodiments, generating a baseline sequence representing the continuously changing portion of a sequence of first or second characteristic indices distributed along a path segment can be achieved as follows: Assume we have an original sequence, for example, a sequence of first characteristic indices representing a certain region on the working path of a turning tool as [10, 11, 10, 100, 12, 13, 14, 5, 15, 16]. First, a local variation index representing the degree of local variation of a data point can be determined based on the data points in the sequence and their neighboring data points. For example, the average absolute difference between each data point and its two preceding and following neighboring data points can be calculated as the local variation index. For data point 100 in the sequence, its neighboring data points are 10 and 12, and its local variation index will exhibit a high value. Second, based on the local variation index, the data points in the sequence are distinguished into trend points representing continuously changing portions or abrupt changes caused by microscopic defects. A threshold for the local variation index can be set; for example, if the local variation index exceeds a certain preset value, the data point is determined to be an abrupt change point; otherwise, it is a trend point. In the above sequence, the local variation indices of data points 100 and 5 may exceed the threshold, thus being identified as abrupt change points, while other data points (10, 11, 10, 12, 13, 14, 15, 16) are identified as trend points. Finally, a baseline sequence is generated based on the distinction between trend points and abrupt change points. The value at the position corresponding to the trend point in the baseline sequence is determined by the value of the trend point in the sequence. For example, 10, 11, 10, 12, 13, 14, 15, and 16 in the original sequence will be directly retained in their corresponding positions in the baseline sequence. The value at the position corresponding to the abrupt change point in the baseline sequence is determined by the value of the trend point adjacent to the abrupt change point in the sequence. For example, for abrupt change point 100, its adjacent trend points could be 10 in front of it and 12 behind it. An interpolation method, such as linear interpolation, can be used to determine the value of abrupt change point 100 in the baseline sequence. If linear interpolation is used, the value at position 100 can be replaced by the average of 10 and 12 (11). For mutation point 5, its neighboring trend points can be 14 and 15, and its value in the baseline sequence can be replaced by the average of 14 and 15 (14.5). In this way, the original sequence [10, 11, 10, 100, 12, 13, 14, 5, 15, 16] can be transformed into a smoother and more representative baseline sequence, such as [10, 11, 10, 11, 12, 13, 14, 14.5, 15, 16], thereby effectively filtering out local mutation information caused by microscopic defects.

[0106] In this embodiment of the invention, the value at the position corresponding to the mutation point in the reference sequence is determined by the value of the trend point adjacent to the mutation point in the sequence, including: determining a first trend point located before the mutation point and a second trend point located after the mutation point in the sequence; and determining the value at the position corresponding to the mutation point in the reference sequence based on the values ​​of the first trend point and the second trend point in the sequence and their respective positions.

[0107] Here, the first trend point refers to a data point in the sequence that precedes a specific abrupt change and is identified as representing the continuous trend of the sequence. It can be determined by using the nearest data point before the abrupt change that meets the trend point criteria. The second trend point refers to a data point in the sequence that follows the specific abrupt change and is also identified as representing the continuous trend of the sequence. It can be determined by using the nearest data point after the abrupt change that meets the trend point criteria. Determining the value of the abrupt change point at its corresponding position in the baseline sequence involves calculating the value that the abrupt change point should have in the smoothed or corrected baseline sequence based on the information from the identified first and second trend points. This can be done using interpolation algorithms, such as linear interpolation, spline interpolation, or weighted averaging. The aim is to eliminate local abrupt changes caused by microscopic defects, allowing the baseline sequence to more accurately reflect the overall continuous trend of the sequence.

[0108] When abrupt changes caused by microscopic defects are identified in the sequence of the first or second characteristic indicators distributed along the path segment, this scheme no longer simply ignores or coarsely processes these points. Instead, it first precisely locates the two nearest trend points before and after the abrupt change in the sequence—the first trend point and the second trend point. Because these trend points represent the continuously changing portion of the sequence, their values ​​provide the "normal" or "expected" value range of the abrupt change without the influence of microscopic defects. Subsequently, based on the specific values ​​of these two trend points in the sequence and their position relative to the abrupt change, this scheme can use a more precise calculation method to determine the value of the abrupt change at its corresponding position in the baseline sequence. This method fully utilizes the local continuity of the sequence. By effectively utilizing the trend points before and after the abrupt change, it can achieve accurate estimation or interpolation of the data at the abrupt change, thereby effectively filtering out local abrupt changes caused by microscopic defects when generating the baseline sequence, allowing the baseline sequence to more accurately represent the continuously changing portion of the sequence. This precise method of generating reference sequences, closely integrated with the steps in the previous scheme, allows for a more accurate reflection of the physical state of the workpiece material when subsequent statistical values ​​representing its spatial continuity are determined as the third characteristic index based on the reference sequence. This third characteristic index can more realistically reflect the actual distribution of complex physical layering structures or microscopic defects. Furthermore, the accuracy of the judgment result is significantly improved when determining the turning physical state of the fuzzy state location based on the third characteristic index. Ultimately, this precise judgment of the workpiece's physical state enables the selection of the corresponding adjustment rule from the preset adjustment rules based on the turning physical state, and the determination of the target spindle speed at different positions on the toolpath. This allows the determined target spindle speed to more accurately adapt to the non-uniform changes in the local material rigidity of the workpiece, effectively avoiding cutting chatter induced by local offsets in system stability boundaries, and significantly improving the stability and machining quality of milling and turning processes.

[0109] In some preferred embodiments, linear interpolation can be used when it is necessary to determine the value at the position corresponding to the mutation point in the baseline sequence. Specifically, suppose a mutation point is located at position P_m in the sequence, with an initial value of V_m. First, the system scans forward to find the nearest trend point before P_m, marking it as the first trend point with position P_1 and value V_1. Next, the system scans backward to find the nearest trend point after P_m, marking it as the second trend point with position P_2 and value V_2. Once the first and second trend points are determined, the estimated value of the mutation point at its corresponding position in the baseline sequence can be calculated using a linear interpolation formula based on their numerical and positional information. For example, the estimated value can be calculated as V_1 + (V_2 - V_1) * ((P_m - P_1) / (P_2 - P_1)). In this way, the values ​​at the mutation points are smoothly transitioned, eliminating the influence of local mutations. This ensures that the generated reference sequence also maintains continuity and smoothness at the mutation points, thus more accurately reflecting the true physical state of the workpiece material.

[0110] The aforementioned technical solution effectively eliminates the interference of local abrupt changes caused by microscopic defects on the determination of sequence continuity by identifying trend points adjacent to the abrupt change point and performing calculations based on the values ​​and positions of these trend points. This allows the generated benchmark sequence to more accurately characterize the continuous physical state of the workpiece material, providing a reliable data foundation for subsequent accurate determination of the workpiece's physical state, thereby improving the accuracy and stability of vibration control in milling and turning processes.

[0111] The step of determining the turning physical state of a newly machined workpiece based on a first feature index and a second feature index in this embodiment of the invention includes: constructing a two-dimensional feature space with the first feature index and the second feature index as coordinate axes, the two-dimensional feature space being used to delineate the corresponding discrimination regions for different turning physical states; forming a feature point in the two-dimensional feature space by combining the first feature index and the second feature index at a position along the working path of the turning tool; and determining the turning physical state of the newly machined workpiece based on the discrimination region in which the feature point falls in the two-dimensional feature space.

[0112] Here, the two-dimensional feature space refers to the mathematical plane formed by the first feature index and the second feature index as two orthogonal coordinate axes. The discrimination region refers to the specific geometric region pre-defined in the two-dimensional feature space for different turning physical states, which is to provide a clear basis for the classification of feature points. The feature point refers to the unique position point in the two-dimensional feature space represented by the values ​​of the first feature index and the second feature index of a position along the working path of the turning tool. Specifically, it can be a coordinate pair composed of (first feature index value, second feature index value), the purpose of which is to present the comprehensive physical characteristics of the current position in an intuitive spatial position form.

[0113] The proposed solution constructs a two-dimensional feature space using the first and second characteristic indices obtained from analyzing the process signals at each position along the turning tool's working path. Within this space, corresponding discrimination regions are pre-defined for different turning physical states, transforming the complex problem of physical state judgment into an intuitive geometric region assignment problem. When determining the turning physical state of a newly machined workpiece, the first and second characteristic indices at a position along the turning tool's working path form a feature point in the two-dimensional feature space. Subsequently, by analyzing the discrimination region where this feature point falls, the turning physical state of the newly machined workpiece at its current position can be determined. This method comprehensively considers the impact and energy characteristics of the process signals, avoiding the limitations and inaccuracies that may exist with single-indicator judgments, and effectively reducing the impact of noise interference on the judgment results. It is precisely this judgment mechanism based on multi-indicator collaborative analysis that makes the identification of the turning physical state of a newly machined workpiece more accurate and reliable. In the overall milling and turning vibration control method, accurate judgment of the turning physical state is a crucial prerequisite for subsequent selection of adjustment rules and determination of the target spindle speed. This solution allows for more precise identification of complex physical layering structures or internal microscopic defects that may exist on the workpiece at different locations. This enables more targeted steps in subsequent operations, such as selecting appropriate adjustment rules from preset rules based on the turning physical state, and determining the target spindle speed at different positions on the toolpath based on these rules and process signals along the tool's working path. This not only improves the effectiveness of spindle speed adjustment but also enhances the adaptability and robustness of the entire milling and turning vibration control method, ultimately helping to avoid chatter and ensure machining quality and efficiency.

[0114] In this two-dimensional feature space, multiple discrimination regions can be pre-defined by analyzing and labeling a large amount of historical machining data. For example, a "stable machining region" can be defined, characterized by low peak factor and root mean square (RMS) values; a "slight vibration region" characterized by a slight increase in either the peak factor or RMS value; a "severe vibration region" characterized by a significant increase in both the peak factor and RMS value; and a "material defect region" characterized by an abnormal abrupt change in the peak factor while the RMS value remains relatively unchanged. Specifically, when the process signal at a position along the working path of the turning tool is collected and analyzed, the peak factor and RMS value corresponding to that position can be calculated. These two values ​​are used as coordinates to determine a feature point in the constructed two-dimensional feature space. For example, if the calculated peak factor is X and the RMS value is Y, then the coordinate point (X, Y) is formed in the two-dimensional feature space. Subsequently, the system can determine which pre-defined discrimination region the feature point falls into. For example, if a feature point falls into the "severe vibration zone," it is determined that the newly machined workpiece is currently under severe vibration; if it falls into the "material defect zone," it is determined that there is a material defect at that location. In this way, the specific physical state of the newly machined workpiece during the turning process can be identified intuitively and accurately, providing a reliable basis for subsequent adjustments to machining parameters.

[0115] The step of determining the target spindle speed corresponding to each position on the working path of the turning tool based on preset adjustment rules in this embodiment of the invention includes: determining the initial target spindle speed based on the process signal characteristics of the working path distribution of the turning tool; during the execution of subsequent turning operations on the new workpiece, collecting real-time process signals reflecting the current interaction state between the turning tool and the new workpiece; determining the characteristics of the current wear state of the turning tool from the collected real-time process signals; setting compensation logic according to the characteristics of the current wear state of the turning tool, and determining the spindle speed compensation amount based on the compensation logic; and correcting the initial target spindle speed according to the spindle speed compensation amount to obtain the target spindle speed.

[0116] The process signal characteristics of the turning tool's working path distribution refer to the signal data acquired during the machining process as the turning tool moves along the predetermined path, reflecting the cutting state and workpiece material response. These can be characterized using cutting force signals, vibration signals, acoustic emission signals, or power signals, and their purpose is to provide basic data for subsequently determining the initial target spindle speed. The characteristics of the turning tool's current wear state refer to parameters extracted from the real-time process signals that quantify or indicate the degree and form of the turning tool's wear. These can be represented by the average cutting force, cutting force fluctuations, specific frequency component energy of the vibration signal, and the amplitude or power trend of the acoustic emission signal. Their purpose is to accurately assess the actual working state of the tool and provide a basis for subsequent compensation logic settings. The compensation logic refers to the rules or algorithms established based on the characteristics of the turning tool's current wear state to calculate the spindle speed compensation amount. It can be implemented using lookup table-based mapping relationships, fuzzy control rules, neural network models, or adaptive control algorithms. Its purpose is to dynamically adjust the spindle speed according to the actual tool wear situation to offset the adverse effects of wear on machining stability and quality.

[0117] In practice, based on the process signal characteristics of the turning tool's working path distribution, the system determines an initial target spindle speed. This initial speed is calculated according to preset adjustment rules (e.g., a mapping relationship between process signals based on physical characteristic information and stable milling spindle speeds), providing a benchmark for subsequent dynamic adjustments. During subsequent turning operations on the newly machined workpiece, the system continuously collects real-time process signals reflecting the current interaction state between the turning tool and the workpiece. These signals, such as cutting forces, vibrations, or acoustic emission signals, directly reflect the actual contact between the tool and the workpiece and the dynamic response of the cutting process. Next, from the collected real-time process signals, the system determines the characteristics of the current wear state of the turning tool. This is achieved by analyzing the real-time signals and extracting characteristic parameters related to the degree and form of tool wear. Subsequently, based on the characteristics of the current wear state of the turning tool, the system sets corresponding compensation logic and determines the spindle speed compensation amount based on this logic. This compensation logic is dynamic and can calculate the required adjustment amount of the spindle speed according to the actual tool wear situation. For example, when increased tool wear is detected, the compensation logic may instruct a reduction in spindle speed to maintain cutting stability; conversely, if wear is minor, it may allow maintaining or slightly increasing the speed to ensure efficiency. Finally, the system algebraically superimposes the calculated spindle speed compensation amount with the previously determined initial target spindle speed to obtain the final target spindle speed. This corrected target spindle speed more accurately adapts to the actual wear state of the tool, ensuring that the spindle speed remains within a range that effectively suppresses vibration and guarantees machining quality and efficiency throughout the machining process. It is precisely because of this closed-loop control of real-time monitoring, evaluation, and dynamic compensation that this solution can overcome the adverse effects of tool wear on machining stability, enabling the initial target spindle speed determined based on preset adjustment rules to adaptively adjust according to actual working conditions. This maintains machining stability and accuracy even when tool wear occurs, significantly improving the adaptability and reliability of turning operations.

[0118] In some preferred embodiments, this application is implemented as follows: First, before machining begins, the system, based on the process signal characteristics of the turning tool's working path distribution, such as by analyzing historical machining data and combining workpiece material properties and tool geometry, uses a preset mapping model or lookup table to determine the initial target spindle speed corresponding to each position on the turning tool's working path under ideal tool conditions. This initial speed can be stored as a path-speed correspondence sequence. During the subsequent turning operations of the newly machined workpiece, the system uses sensors installed on the machine tool spindle or tool holder, such as piezoelectric force sensors or accelerometers, to collect real-time process signals reflecting the current interaction state between the turning tool and the newly machined workpiece, such as cutting force signals or vibration signals. These signals are continuously acquired at a high sampling rate. Next, a signal processing module determines the characteristics of the current wear state of the turning tool from the collected real-time process signals. For example, by performing mean and variance analysis on the real-time cutting force signal, or by performing spectral analysis on the vibration signal, characteristic parameters related to tool wear are extracted, such as the upward trend of the average cutting force and the increase of vibration energy within a specific frequency range. These feature parameters can be input into a pre-trained tool wear condition recognition model, which can be a support vector machine or a neural network, to determine the current wear level of the tool, such as slight wear, moderate wear, or severe wear. Subsequently, based on the identified characteristics of the current wear state of the turning tool, the system sets compensation logic. This compensation logic can be a rule base based on fuzzy reasoning; for example, if the tool is in a moderate wear state and the average cutting force exceeds a certain threshold, a negative spindle speed compensation amount is set; if the tool is in a severe wear state, a larger negative compensation amount is set. Based on this compensation logic, the system calculates the specific spindle speed compensation amount. Finally, the system algebraically superimposes the calculated spindle speed compensation amount with the previously determined initial target spindle speed to obtain the final target spindle speed. For example, if the initial target spindle speed is 1000 rpm and the compensation amount is -50 rpm, then the final target spindle speed will be corrected to 950 rpm. This corrected target spindle speed is then sent to the machine tool's CNC system for real-time spindle speed control to ensure the stability and quality of the machining process.

[0119] Specifically, Figure 3 A schematic diagram of a milling and turning vibration control system according to an embodiment of the present invention is shown. This system is used to execute a milling and turning vibration control method and includes:

[0120] The feature construction module is used to collect process signals associated with the position of the turning tool in the historical turning process of the machined workpiece, and to establish the physical feature information of the machined workpiece under the turning process based on the process signals.

[0121] The path construction module is used to match the corresponding turning tool working path for the new workpiece based on physical feature information, and to associate the corresponding process signal according to the turning tool working path.

[0122] The spindle speed determination module is used to match the process signal to determine the target spindle speed at each position on the working path of the turning tool.

[0123] The machine tool control module is used to control the newly machined workpiece to perform turning operations on the machine tool along the working path of the turning tool based on the target spindle speed.

[0124] The proposed solution first uses a feature construction module to collect process signals associated with the turning tool position from the historical turning operations of the workpiece, and then establishes physical characteristic information of the workpiece during the turning process based on these signals. This physical characteristic information is a digital representation of the mechanical response and material properties exhibited by the workpiece during actual machining, laying the data foundation for subsequent intelligent decision-making. Based on this, a path construction module receives this physical characteristic information and uses it to match the most suitable turning tool working path for the new workpiece. Simultaneously, this module can associate the selected tool working path with the process signals that may be generated along that path, which is equivalent to a prediction of future machining states. Furthermore, the speed determination module uses the process signals associated by the path construction module to accurately match the target spindle speed corresponding to each position on the turning tool working path. This step is crucial for vibration control, ensuring that the spindle operates at the most stable speed at every point along the tool's path, thereby effectively avoiding resonance. Finally, the machine tool control module converts the target spindle speed output by the speed determination module into instructions that the machine tool can recognize and execute, controlling the newly machined workpiece to perform turning operations along the working path of the turning tool. Through refined management and dynamic adjustment of the machining process, this system significantly improves the stability of milling and turning operations, thereby effectively suppressing machining vibrations, improving machining quality, and extending tool life.

[0125] The system in this embodiment of the invention organically integrates key aspects such as data acquisition, feature extraction, path planning, spindle speed optimization, and machine tool control, enabling the previously abstract vibration control strategy to be implemented in a real-world machining environment. This solves the problem that relying solely on methods makes it difficult to achieve real-time, precise control of complex milling and turning processes. Through close collaboration and information flow between modules, the system can intelligently plan machining parameters and dynamically adjust the spindle speed based on historical data and workpiece characteristics, thereby significantly improving the stability of the machining process, effectively suppressing cutting vibration, improving the surface quality of the workpiece, reducing tool wear, and increasing overall machining efficiency.

[0126] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A vibration control method for milling and turning machining, characterized in that, include: Collect process signals associated with the position of the turning tool during the historical turning process of the workpiece, and establish physical characteristic information of the workpiece under the turning process based on the process signals; Based on the physical feature information, a corresponding turning tool working path is matched for the newly processed workpiece, and the corresponding process signal is associated with the turning tool working path; Process signals refer to physical or state quantities acquired in real time during the turning process, corresponding to the specific position of the turning tool on the workpiece. Physical characteristic information refers to a set of data or models that can characterize the mechanical properties or state of a machined workpiece under turning action by analyzing, processing and modeling process signals. Based on the process signal, the target spindle speed corresponding to each position on the working path of the turning tool is matched; Based on the geometry, material type, and preset machining path of the new workpiece, search or derive the data or model that is closest to the physical response or state that the new workpiece may produce on a specific turning tool working path from the established physical feature information. Based on the target spindle speed, the newly machined workpiece is controlled to perform turning operations in the machine tool along the working path of the turning tool; The process of matching the target spindle speed at each position on the working path of the turning tool based on the process signal includes: Preset adjustment rules are set based on the mapping relationship between process signals in physical feature information and the stable spindle speed of milling. The target spindle speed corresponding to each position on the working path of the turning tool is determined based on the preset adjustment rules.

2. The vibration control method for milling and turning machining according to claim 1, characterized in that, The determination of the target spindle speed corresponding to each position on the working path of the turning tool based on the preset adjustment rules includes: The process signals on the working path of the turning tool are analyzed to obtain a first characteristic index and a second characteristic index. The first characteristic index represents the impact characteristics of the process signals on the working path of the turning tool, and the second characteristic index represents the energy characteristics of the process signals on the working path of the turning tool. The turning physical state of the newly processed workpiece is determined based on the first characteristic index and the second characteristic index; Select the corresponding adjustment rule from the preset adjustment rules based on the physical state of the turning process; The target spindle speed at different positions on the toolpath is determined based on the corresponding adjustment rules and the process signals on the working path of the turning tool.

3. The vibration control method for milling and turning machining according to claim 2, characterized in that, The step of determining the turning physical state of the newly machined workpiece based on the first characteristic index and the second characteristic index includes: When it is determined that the combination relationship between the first feature index and the second feature index satisfies the preset fuzzy physical state, the position on the working path of the turning tool is determined as the fuzzy state position. Extract a path segment containing the fuzzy state position on the working path of the turning tool; The first and second feature indicators distributed on the path segment are analyzed, and a third indicator feature representing the fuzzy physical state within the path segment is obtained. The turning physical state at the location of the fuzzy state is determined based on the third indicator feature. The turning physical state includes: composite physical layered structure and microscopic defects inside the workpiece material.

4. The vibration control method for milling and turning machining according to claim 3, characterized in that, The step of parsing the first and second feature indicators distributed on the path segment and obtaining the third indicator feature representing the fuzzy physical state within the path segment includes: For the sequence of the first feature index or the second feature index distributed on the path segment, a reference sequence is generated to characterize the continuously changing part in the sequence, and the reference sequence has filtered out local mutation information caused by the micro defects; Based on the benchmark sequence, a statistical value representing the spatial continuity of the benchmark sequence is determined as the third feature index.

5. The vibration control method for milling and turning machining according to claim 4, characterized in that, Generating a reference sequence characterizing the continuously changing portion of the sequence based on the sequence of the first or second feature indicators distributed along the path segment includes: Based on the data points in the sequence and the neighboring data points of the data points in the sequence, a local variation index is determined to represent the degree of local variation of the data points; Based on the local variation index, the data points in the sequence are distinguished into trend points representing the continuously changing part or abrupt change points caused by the microscopic defects; The baseline sequence is generated by differentiating the results based on the trend point and the mutation point, wherein the value at the position corresponding to the trend point in the baseline sequence is determined by the value of the trend point in the sequence, and the value at the position corresponding to the mutation point in the baseline sequence is determined by the value of the trend point adjacent to the mutation point in the sequence.

6. The vibration control method for milling and turning machining according to claim 5, characterized in that, The value at the position corresponding to the mutation point in the baseline sequence is determined by the values ​​of the trend points adjacent to the mutation point in the sequence, including: In the sequence, identify a first trend point located before the mutation point and a second trend point located after the mutation point; Based on the values ​​and positions of the first trend point and the second trend point in the sequence, the value of the mutation point at the corresponding position in the baseline sequence is determined.

7. The vibration control method for milling and turning machining according to claim 2, characterized in that, The step of determining the turning physical state of the newly machined workpiece based on the first characteristic index and the second characteristic index includes: A two-dimensional feature space is constructed using the first feature index and the second feature index as coordinate axes. The two-dimensional feature space is used to delineate the corresponding discrimination regions for different turning physical states. The first feature index and the second feature index at a position along the working path of the turning tool are combined to form a feature point located in the two-dimensional feature space. The turning physical state of the newly processed workpiece is determined based on the discrimination region in which the feature points fall in the two-dimensional feature space.

8. The vibration control method for milling and turning machining according to claim 1, characterized in that, The determination of the target spindle speed corresponding to each position on the working path of the turning tool based on the preset adjustment rules includes: The initial target spindle speed is determined based on the process signal characteristics of the working path distribution of the turning tool. During the subsequent turning operations of the newly machined workpiece, real-time process signals reflecting the current interaction state between the turning tool and the newly machined workpiece are collected. The characteristics of the current wear state of the turning tool are determined from the collected real-time process signals; The compensation logic is set according to the characteristics of the current wear state of the turning tool, and the spindle speed compensation amount is determined based on the compensation logic; The target spindle speed is obtained by correcting the initial target spindle speed based on the spindle speed compensation amount.

9. A vibration control system for milling and turning, used to execute the vibration control method for milling and turning as described in any one of claims 1-8, characterized in that, The system includes: The feature construction module is used to collect process signals associated with the position of the turning tool in the historical turning process of the processed workpiece, and to establish the physical feature information of the processed workpiece under the turning process based on the process signals. The path construction module is used to match the corresponding turning tool working path for the new workpiece based on the physical feature information, and to associate the corresponding process signal according to the turning tool working path. The rotational speed determination module is used to match the target spindle speed corresponding to each position on the working path of the turning tool based on the process signal; The machine tool control module is used to control the newly machined workpiece to perform turning operations in the machine tool along the working path of the turning tool based on the target spindle speed.

Images